Vol. 2, No. 3, June, 1963
SOrES
0.10
09
.08
.07
.OG
.-d
; .OB
I
u
.04
.03
.02
.01
10
20
30
40
50
60
70
Time, hr.
Fig. 3.-Decomposition
of monochlorodiborane in nitrogen at 35'.
a 33 ft. X 0.25 in. stainless steel tube, containing 60-80 mesh firebrick coated with 20% (wt.) silicone oil (Dow Corning No. 703). With the column maintained a t 0" and a 100 ml./min. He flow, BzH5C1emerged as a pure fraction a t 22.5 min. By using numerous samples of the mixture, quantities were selectively trapped in a section of the vacuum line a t liquid nitrogen temperature. By infrared analysis, no traces of BCla, or BHClz were detected in the collected fractions. Anal. Calcd. for BzHfiCl: B, 34.8; H, 8.1; C1,57.1. Found: B,34.0; H,8.4; C1,56.7. Based on P Vir measurements a t 25", a molecular weight corresponding to 60.5 was found (calculated = 62.1). The infrared spectrum for monochlorodiborane a t 8.8 mm. is given in Fig. 1. The cell path length was 10 cm. and the temperature was 25". Preparation of BHCl2.-Preheated streams of Ht and BC13 (in a 1 1 mole ratio) were passed through a quartz tube maintained a t 900". With a retention time of 1.7 sec., a 177, conversion of BC13 occurred to give a BHClz yield of 97.3y0, No traces of B2HsC1, BzHs, or BzCL were present in the gaseous product mixture. By condensing the BHClz and BCI3 a t - 11l0, excess Hz and HCl were removed from the mixture. Dichloroborane samples then were separated from BCl3 by employing a preparative, low-temperature, gas chromatography column.8 Anal. Calcd. for BHClz: B, 13.1; H, 1.2; C1,85.7. Found: B, 12.6; H , 1.0; c1, 87.1. An infrared spectrum of the compound is shown in Fig. 2. The tracing was obtained a t a total gas pressure of 16 mm. in a 5.0-cm. path length a t -10". By placing 0.6 mm. of BC13 in an identical 5.0-cm. reference cell a t 30", absorption due t o this impurity was "blanked" from the spectrum. The resulting spectrum agrees well with the absorption peaks reported by other investigators .6, QJO Conversion of BHClz to BzH,Cl.-A solution (300 ml.) containing 3.93% (mole) BHClz in liquid BC18 reached an equilibrium vapor pressure of 599.6 mm. a t 0.5' in a 2-1. vacuum(9) H G Nadeau and D. M. Oaks, Jr , Anal Chem , 32, 1481 (1960) (10) T Onak, H.L. Landesman, and I. Shapiro, J Phys Chem , 62, 1605 (1955)
657
jacketed, glass chamber. Samples of the vapor phase, which contained 21.20% (mole) BHC12, were raised slowly to 40' under autogenous pressure while gas chromatograms and infrared spectra were recorded. The first analysis, a t lo", showed that 0.91% BzH6Cl(but no BzHB)was present in the mixture. A concentration of 2.54% BzH6C1 was reached before traces of BZH8 appeared. Reaction of BzH5Cl with Excess BC13.-A 0.214-mmole sample of BzH~CI was condensed in a vacuum-line trap with 50.23 mmoles of BCll and allowed to stand a t 30' for 4 hr. (liquid BC13 was present). The mixture was vaporized completely into a large known volume and analyzed by infrared, gas chromatography, and gasometry. The BzH~CIhad diminished t o 0.132 mmole while 0.391 mmole of BHClz had formed in the mixture. S o quantities of BZH8j HCI, or Hz were detected. After 1 week a t 20', an identical mixture showed that all the B2HsClhad reacted to form BHC12. Disproportionation of BzH&l.-Under a pressure of 760 mm ., a gas mixture containing 9.1% B~HsCI in nitrogen was maintained a t 35' in a 2-1. glass chamber. Over a 70-hr. period, small samples were removed for analyses by gas chromatography, gasometry, and infrared spectrometry. The results are shown in Fig. 3. Reaction of B&CI with B203.-A 1.0-g. (14.36 mmoles) was placed in the U section of a cold sample of - 100 mesh B z O ~ trap attached to the side of an infrared gas cell. A 0.306-mmole sample of BzH5Cl was admitted to the gas cell and the spectrum recorded. Liquid nitrogen was placed around a remote tip of the U-trap, causing all the BzH5Cl to condense there. As the tip was warmed to 25", the gas returning to the cell passed through the B ~ 0 3a second time. The infrared spectrum then showed only B&8 (0.235 mmole) to be present in the gas cell. After evacuating the BZH8, the Bz03 was heated to 100' as an infrared scan was recorded. A spectrum of BC13 then was observed. After evacuating the BCI3, analysis of the resulting B203 showed no C1 to be present. Reaction of B2H5Cl with Excess HC1.-To a 0.415-mmole sample of BzHsCl contained in an infrared gas cell, a 5.302-mmole sample of HC1 was introduced a t 25". An infrared scan of the 1500 to 1650 cm. -l region was begun and showed that B&T6C1had immediately disappeared. Gas chromatographic and gasometric analysis of the mixture showed 2.052 mmoles of Hz and 0.851 mmole of BCl3 to be present in the remaining HCl.
Acknowledgment.-This work was performed under Contract No. AF33(600)-35780 with the United States Air Force, Aeronautical Systems Division, Manufacturing Technology Laboratory, Wright-Patterson Air Force Base, Ohio.
CONTRIBUTION FROM THE NATIONAL BUREAU OF STANDARDS WASHINGTON, D. c.,AND THE DRPARTMENT O F CHEMISTRY O F THE UKIVERSITY O F WISCONSIN, MADISON, WISCONSIN
The Nature of the Compound Obtained from Aqueous Cesium Chloride Solution and Hydrogen Chloride BY ARTHURG. MAKIA N D ROBERT U'EST
Received Janaiauy 17, 1963
In 1957 one of us reported the preparation of a new compound from aqueous cesium chloride solution saturated with hydrogen chloride.' On the basis of partial (1) K . B'rst, J . A m . Chem. SOC.79,4568 (1957).
658 NOTES
Inorganic Chemistry
analysis this substance was formulated as CsHC12. Recently ValleC and McDaniel have shown by means of more complete analyses that the correct formula for the compound is C S C ~ 4HCI.3/4H20.2 .~, The same workers also have prepared the compound CsHClz and established that i t is stable only a t very low temperatures or high hydrogen chloride pressures. These findings make it seem unlikely that the compound obtained from aqueous solution contains the hydrogen dichloride anion HCla-, but leave unsettled the question of the actual nature of the substance. In this paper infrared spectral evidence will be presented which indicates that the compound is probably a mixed oxonium salt, 4Cs+C1-.3H3O+CI-. Experimental The compound CsCl .3/;HCI. 3/4Hz0was prepared by the originally published procedure.' The deuterium analog was obtained in an analogous manner from DCl in 977, D20. Infrared spectral examination of the water liberated upon decomposition of the deuterium compound indicated that deuteration was more than 95% complete. Infrared spectra were determined for both substances as mulls in Kujol, using a Beckman IR-7double-beam grating-prism spectrometer which was flushed with dry air. Typical spectra are shown in Fig. 1.
Results and Discussion The spectrum consists of a series of broad bands, all shifted to lower frequency in the deuterium compound. The positions of absorption all correspond to bands previously reported for H30+ and D30+ compounds studied a t low temperatures, and the over-all agreement of the spectra with those of H 3 0 + and D30+ is very g ~ o d . ~The , ~ spectra are quite unlike those for typical ionic substances containing hydrate water, and bands characteristic of HCI and DC1 are also a b ~ e n t . ~ The spectrum of (CH3)&+HCl~-has been studied by "addington, who attributed bands a t 1566 and 1180 cm.-' to the HC12- ion.6 These also are absent in the spec( 2 ) R . E. Vallee and D. H. Mcnaniel, J . A m . Chem. Soc., 84, 3412 (1962). (3) C. C. Ferris0 and D. F. Hornig, ibid., 76, 4113 (1953); J . Chem. Phys., 23, 1464 ( 1 9 5 5 ) . (4) D. E. Bethell and N. Sheppard, ibid.,23, 1421 (1953). ( 5 ) D . F. Hornig and G . I,. Hiebert, i b i d . , 27, 734 (1957). (6) T. C. Waddington, J . Chenz. Soc., 1708 (1988).
4000
3000
2000
1000
CM-' Fig. 1.-Infrared spectra of 4CsC1.3HaOCl (solid line) and 4CsC1.3D30Cl (dashed line) as mulls in Sujol. Kujol bands have been deleted.
trum of the cesium compound. Of the likely absorbing species, only oxonium ion seems to be present, and hence the spectral evidence suggests that the compound is actually 4CsCl. 3H30C1. This structure is entirely reasonable, for it is known that HC1 transfers a proton to water in its solid monohydrate, forming HaO+C1-.3 However, the stabilization of the oxonium ion by cesium chloride in 4CsC1-3H80C1is remarkable. Some variability was noted in the spectra of different mulls of the same sample of cesium compound. The two bands a t about 2650 and 2500 cm.-' occasionally appear as a single broad band centered near 2630 cm.-l. The intensity and position of the 2170 cm.-' band is somewhat variable, and the intensity of the 785 cm.-' band changed markedly. Some of the variability may be due t o differences in particle size and hence in the amount of scattered light. However, the results also suggest preferred orientation of crystallines in certain mulls, and variability in absorption depending on crystal orientation. Some experiments to determine the presence of HCIZ- ion in water also have been carried out, with negative results. It was found that the presence of up to 8 -If lithium chloride does not alter the infrared or Raman spectrum of hydrochloric acid solutions of various concentrations.
